March 9, 2011

How to optimize marketing for serving customers across multiple websites

The fun part about creating websites for a business is that you are not automatically limited to having one website. However, many businesses do not realize that with some planning, their analytics solution can be a business intelligence tool that permits better marketing decisions and better chances to be profitable.

First a brief explanation -- why track, even if the websites are not an e-commerce or a blog? A web analytics tool can indicate what referral sources contributed to the site, allowing businesses to better adjust their marketing.

An analytics solution works by attributing the last website a visitor came from as the referral source. In a sequence where the visitor can come from one website to a second, then to a third website tagged with a web analytics code, the second website is considered as the referral source in the analytics reports.

This scenario, which often occurs when running multiple sites, becomes a distorted view of traffic source contributors - you want to treat some visitor sessions as one visit across domains if those visitors are viewing pages in one sitting. The advantage of a multiple site tagging leads to better understanding of what attracts visitors and can affect website redesign decisions, product and service offerings, and even the business model itself.

To implement a multiple site tag, an analytics code is typically modified in the Javascript code. This step sounds technical, but it is not extremely so. Google Analytics, for example, employs a _pagetracker javascript callout in the code. The bold font below highlights where the pageTracker callout appears: The setDomainName addresses domain recognition in the Google Analytics cookie, permitting the desired recognition across domains, while the setAllowlinker allows for the Google Analytics cookie recognition through the link function between the domains.

Google Analytics code with multiple domain modification

In late 2009, Google Analytics introduced an asynchronous version of the code. The asynchronous version was introduced to address page loading speed, but has become now its latest version - you can learn the difference at the Google code site here. If you’ve added Google Analytics to your site since 2009, chances are you are using the asynchronous version. The asynchronous version of the modified code as such below. It similar to one above, except for the _Gaq.push script that is used.

Asynchronous Google Analytics code with multiple domain revision

After the Google Analytics code is modified, the next step is modifying links to the second will then add an on-click _link function to all the links between the tagged websites. For example, a HMTL link code that takes the visitor to the second site....

A link for a second domain

...would look like the one below with a pageTracker link function called by the on-click event:

Second domain link modified for multiple domain visitor tracking

This is the standard version. The asynchronous version looks like this below (Note -- gap between link on onclick is due to blog format: No actual space is required):

Second domain link with multiple domain tracking

For submitting forms between domains, GA uses a different function at the link. It would like the following:

Post version of second domain link

This example is typically used when data is submitted across domain, such as entry windows on a contact page.

Other analytics solutions have similar arrangements. For example Piwik offers a similar set up for its analytics solution. Piwik users can learn more about setting separate websites, including subdomains here.

The point is that with good measurement, great service for all your website layout arrangements are possible. For your business, work with your design team to make sure the tagging reflects the way your business operates online.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram